Log of the #fcrepo channel on chat.freenode.net

Using timezone: Eastern Standard Time
* github-ff joins05:25
[fcrepo4] acoburn pushed 1 new commit to master: https://git.io/vrrQi
fcrepo4/master 8daa68d Michael Durbin: Updated to code to gracefully handle failure to migrate property. (#1039)...
* github-ff leaves
* travis-ci joins05:46
fcrepo4/fcrepo4#4479 (master - 8daa68d : Michael Durbin): The build passed.05:47
Change view : https://github.com/fcrepo4/fcrepo4/compare/04d534da3b44...8daa68da2b9c
Build details : https://travis-ci.org/fcrepo4/fcrepo4/builds/132505403
* travis-ci leaves
* jgpawletko leaves07:00
* jcoyne joins07:37
* jcoyne leaves07:46
* jcoyne joins07:49
* peichman joins07:59
* peichman leaves08:01
* dwilcox joins08:32
* arebenji joins08:49
* bseeger joins09:04
* whikloj joins09:07
* bseeger leaves09:09
* mikeAtUVa joins09:15
* dwilcox leaves09:27
<mikeAtUVa>So... it seems Sufia (a popular fedora 4-based hydra head) uses the same predicate for it's internal access controls as our webac implementation.09:59
It would seem webac (as currently implemented) is incompatible with Sufia.
* whikloj leaves
* whikloj_ joins
* bseeger joins10:01
<jcoyne>mikeAtUVa: the idea was to make them interoperable.10:05
mikeAtUVa: So can you explain how they are incompatible?
Sufia wants to write access controls that FC4 could enforce for other consumers.10:06
And it's not something that is Sufia specific. All Hydra 10+ is using the same access controls.10:07
That's the point of WebAC, right? So that we all use a common ontology.10:08
* dwilcox joins10:09
<mikeAtUVa>jcoyne: at this point, fedora's webac has some pretty specific requirements, not all of which are met by the instance of sufia a coworker is using to update our shared fedora 4 repository.10:10
<jcoyne>mikeAtUVa: Can you document those and pass them on?10:11
<mikeAtUVa>jcoyne: if the goal is interoperability, I'm willing to spend a little time identifying the current gaps
jcoyne: sure thing!
<jcoyne>mikeAtUVa: At present, I don't think there's any hope of running Hydra 10 with enforcement of WebAC turned on in Fcrepo
mikeAtUVa: is she testing Sufia 6 or the master branch?10:12
And a released version of Fcrepo or master?10:13
<mikeAtUVa>jcoyne: I'm pretty far on the perifery... they're using this https://github.com/uvalib/Libra2 and we've been testing against a bleeding edge fedora (mainly because there have been a couple webac bugs I needed to fix before discovering this)
jcoyne: we have time... they have a setup that works but are working with me in paralelle to work through issues with webac and fedora.10:14
<jcoyne>Yeah, they are on the master branch.
mikeAtUVa: Excellent.10:15
<mikeAtUVa>jcoyne: I want these two to play together by August..
<jcoyne>mikeAtUVa: we could use your help. I know that acoburn, escowles and stephano may be working on that issue too, so you might want to chat with them.
<mikeAtUVa>jcoyne: what method of collaboration would be the best way to help?10:18
jcoyne: github ticket? Jira? IRC?
<jcoyne>For either a github ticket here: https://github.com/projecthydra/hydra-head/issues/new10:19
or put it on the FCRepo jira and send me a link
I suspect there is some hydra specific work (e.g. send current user/groups in a header)10:20
<mikeAtUVa>jcoyne: I believe it... when I figure out what will work (if I don't discover another confounding bug in the meantime) we can determine where the changes need to occur.10:21
* kefo joins10:25
* dwilcox leaves10:33
* bseeger leaves11:06
* jcoyne leaves11:22
* dwilcox joins11:28
* barmintor joins11:29
dwilcox: I remember you talking about a consultancy who had dome sone Fedora work at CalTech. Can you remind me who they were?11:30
<dwilcox>barmintor: Hmm, that may have been IIA (http://iiaweb.com)?11:43
* kefo leaves11:44
<barmintor>dwilcox: maybe!11:50
dwilcox: trying to drum up two more consultancies.
<dwilcox>barmintor: Was it Fedora consulting rather than Hydra or Islandora?
* bseeger joins
<barmintor>dwilcox: I think it was a group that had work ed with Blacklight and was looking into Fedora projects11:51
dwilcox: but I'm hazy
<dwilcox>barmintor: Hmm, yes that sounds familiar. I will try to recall who I was talking about :)11:52
<barmintor>:P
<dwilcox>barmintor: Oh, it may have been Design for Context: http://www.designforcontext.com. They do a lot of linked data stuff and are interested in Fedora but haven't really dug into it yet11:53
<barmintor>hmm.11:54
* jcoyne joins11:55
* bseeger leaves12:15
<arebenji>Hi!12:20
When I add fcrepo-camel-toolbox12:21
I fail to install the fature called fcrepo-service-activemq
I wonder if I've missed anything, is it okay to paste my script? it's 5 lines12:22
<barmintor>arebenji: can I recommend gist: https://gist.github.com/12:34
<arebenji>https://gist.github.com/br2490/bd5eabeb75cadf0f688257272e915b2812:35
lo siento13:25
please ignore my rambling :)13:26
* acoburn joins13:28
* bseeger joins13:44
* ajs6f joins
* bseeger leaves14:22
* acoburn leaves
* bseeger joins14:25
* acoburn joins14:32
<barmintor>I know not all of y'all do twitter but Sara Jeong livetweeting GoogleVsOracle right now is amazing14:42
<ajs6f>Is that the continuing legal saga?
<barmintor>yes
apparently the jury is not able to open the Android and JSE source code submitted as evidence in the court-provided laptop14:43
and the lawyers are hurriedly writing a memo of instruction about how the jury can interact with the tech support staff
<ajs6f>What possible good would it do for the jury to examine source code?14:44
<barmintor>the case is fundamentally about whether Andriod is an infringement on JSE code, apparently14:45
<bseeger>that's insane to have them looking at source code
<barmintor>the Oracle attorney is demanding to know why they didn't submit a printed version of the Android source code
<ajs6f>I can explain that: because you, Oracle att'y, are an idiot.14:46
<barmintor>"The problem is that the files are hidden inside a lot of folders"
"You open that up, there's folder. You open that, there's another folder. You open that, there's four more folders."14:47
<ajs6f>The problem is that the notion of intellectual property is debatably coherent or useful
<barmintor>(those are either the jury or the tech support, unclear)
<ajs6f>Remember when we had to warn people to be careful to use only a Sun JVM for certain operations?14:48
<barmintor>I want to yell "If you think the core Java APIs are bad go look at MODE" but they won't hear me
<ajs6f>I want to go learn Yesod.
<barmintor>I have never seen a high profile software case be argued or tried in a way that makes any sense at all as a developer. Even SCOvLinux was basically "shut up it would do too much damage"14:51
<ajs6f>No, and you never will. Some folks argue that partly that's because software developers do such a rotten job at making their work understood, and there's some truth in that. But the larger truth is that our legal system does a poor job of dealing with expertise.14:52
Engineering expertise, basic science expertise, financial expertise…14:53
{back off soapbox}14:54
<bseeger>Nothing highlights what you're saying better then this "look at the source code" fiasco.
<ajs6f>barmintor: What are _you_ going to write your Fedora reimpl in? Not Java, I hope. I wonder if Scala and Clojure have any exposure to this silliness.14:55
* acoburn leaves14:56
<barmintor>ajs6f: indeed, I was saying this week that Oracle is yanking the rug out from under my long-held defense of Java as an appropriate platform
<ajs6f>barmintor: It's time for Fedora to dev its own language: Fedo. A program in Fedo will consist entirely of complexly nested LDP containers, with semantics attached by means of type assertions. A REPOSITORY WILL BE ITS OWN SOURCE CODE.14:58
<barmintor>Hmm.
<ajs6f>_This_ is what we need to work on at OR.
* bseeger leaves15:01
<ajs6f>ruebot: She missed out on the the Macho Man. Still pretty awesome: https://imgur.com/gallery/pdfTtY8/new15:05
* acoburn joins15:07
* jcoyne leaves15:36
* dwilcox leaves15:41
* dwilcox joins15:44
<barmintor>does anyone remember if you could ingest FOXML with a managed datastream and provide its content in an xmlContent element? sigh.
<mikeAtUVa>barmintor: I think you could.15:45
<ajs6f>barmintor: agree with mikeAtUVa. I vaguely remember doing just that on occasion.15:46
* barmintor grumbles towards trying it15:49
this hydra thing needs to create a bunch of objects, and the piecemeal creation is too slow
thanks mikeAtUVa ajs6f15:50
<mikeAtUVa>barmintor: if it doesn't work just base64 encode it;)
<ajs6f>barmintor: We make dreams happen. As long as they are exactly one kind of dream. The dream of using Fedora Commons software to do stragithforward tasks.15:51
<barmintor>but then I have to have my foxml serialize inspect the content to deviced what wrapping element to use :(
ajs6f: [redacted]
* dwilcox_ joins15:52
<barmintor>deviced!15:53
* dwilcox leaves
<mikeAtUVa>barmintor: I just tested it... it "works" in that the object is ingested, it doens't work in that it just treats it as an X datastream. :(15:55
* dwilcox_ leaves15:56
* dwilcox joins16:00
* dwilcox_ joins16:05
* dwilcox leaves
* jcoyne joins16:23
* acoburn leaves16:25
* arebenji leaves16:26
* dwilcox_ leaves16:32
* dwilcox joins16:34
* dwilcox_ joins16:44
* dwilcox leaves16:47
* whikloj leaves17:04
<barmintor>mikeAtUVa: dangit17:07
* dwilcox_ leaves17:16
* jcoyne leaves17:38
* dwilcox joins17:48
* jcoyne joins17:58
* dwilcox leaves17:59
* jcoyne leaves18:30
* jcoyne joins
* jcoyne leaves
* jcoyne joins18:31
* jcoyne leaves19:16
* jcoyne joins19:27
* jcoyne leaves19:41
* dwilcox joins20:09
* dwilcox leaves20:20
* dwilcox joins20:21
* dwilcox leaves20:54
* mikeAtUVa leaves22:47